When Anthropic’s chief financial officer sat down with prospective investors in August, the questions were not about whether the company would go public. They were about how big the offering would be. The answer, according to people familiar with the discussions, is very big: Anthropic is preparing to publish its S-1 filing as soon as next week, and the company is aiming to raise more than the $86.2 billion that SpaceX collected in its June initial public offering, a record for a single listing.
The figures circulating among investors explain the ambition. Anthropic’s annualized revenue has climbed to roughly $65 billion, up from about $9 billion at the end of last year, according to people who have seen the numbers, a surge driven by demand for the Claude models from businesses that run them on everything from customer service to software writing. The company submitted its registration statement confidentially to the Securities and Exchange Commission on June 1, and investors have valued it at close to $1 trillion in recent funding rounds.
The speed of the timetable is itself a statement. Anthropic’s chief rival, OpenAI, has told investors it is in no hurry, pointing to a listing more likely in 2027, while Anthropic has moved through the confidential filing, the investor meetings and the drafting of a public prospectus in a matter of months. Both companies now sit in the SEC’s review pipeline at nearly the same moment, and bankers say the first to list will set the valuation anchor for the other: the multiple the market assigns to revenue, growth and compute costs will become the yardstick for every frontier lab that follows.
The race reflects how much money the industry needs. Anthropic’s revenue is growing faster than almost any company in history, but so are its bills, and the data centers that train and run its models consume capital on a scale that private investors alone struggle to supply. A listing would give the company a public currency to fund that buildout, and a record raise would put its war chest on the same footing as OpenAI’s.
The prospectus will settle questions that have been circulating among investors for months: how the founders intend to keep control of a company valued near $1 trillion, how large the stakes held by Amazon and Google have grown, and what obligations accompany the safety commitments Anthropic has made to regulators in Washington, Brussels and London. The document is expected to run several hundred pages, and its disclosures on compute spending and the cost of serving Claude will be read as carefully as its revenue figures.
The company arrives at the public market with a story that is both an advantage and a complication. Anthropic was founded in 2021 by Dario and Daniela Amodei and other former OpenAI researchers who broke away over questions of safety, and it has built its reputation on the claim that the most powerful models need careful restraint. Amazon and Google are among its largest backers, and Claude has become a standard choice for enterprises that want an AI assistant with a reputation for reliability rather than spectacle.
The harder sell may be reconciling the two stories the company tells. Its public statements about the dangers of advanced AI helped it stand apart from rivals and attracted the researchers and regulators who take those warnings seriously. A prospectus, by contrast, must promise durable growth, widening margins and a business that compounds for decades, and skeptics will ask whether the caution that defines Anthropic’s brand survives contact with the demands of public shareholders.
Regulators are the other audience. Authorities in the United States and Europe are still deciding how much oversight the most powerful models will face, and Anthropic’s own warnings about catastrophic risk have been cited in those debates. The company will have to answer questions about how its models are tested, who can use them and what happens when they fail, in public filings that carry legal weight rather than in blog posts that do not.
The market’s appetite will be tested as much as the company’s disclosures. A valuation near $1 trillion would place Anthropic among the most valuable publicly traded companies in the world on its first day, a position no technology debut has reached before, and the size of the raise implies demand from sovereign funds, index investors and retail buyers who have never had a direct stake in a frontier AI lab. Underwriters are said to be confident the demand is there; the history of large technology listings suggests the pricing will be a negotiation rather than an auction.
None of this answers the question the company’s founders have spent five years posing about artificial intelligence itself: whether the technology’s promise justifies its risks. The public markets will now deliver their own verdict, pricing those risks in dollars for the first time. Anthropic’s founders argued that AI should be built carefully. The listing will test whether the careful company can also be the fastest, and what the market charges it for trying.
